Slashdot Mirror


AT&T To Match Google Fiber In Kansas City, Charge More If You Want Privacy

An anonymous reader writes: When Google Fiber started bringing gigabit internet to cities around the U.S., we wondered how the incumbent ISPs would respond. Now we know: AT&T has announced they will match Google Fiber's gigabit offerings in Kansas City. Of course, there are some caveats. First, AT&T's rollout may stop as it fights the Obama administration over net neutrality. Not that it would be a nationwide rollout anyway: "AT&T does not plan to offer the ultra-fast Internet lines to every home in the market. Rather, he said the company would calculate where demand is strongest and the investment in stringing new cables promised a decent return."

There are also some interesting pricing concerns. The company plans to charge $70/month for gigabit service, but that's a subsidized price. Subsidized by what, you ask? Your privacy. AT&T says if you want to opt out of letting them track your browsing history, you'll have to pay $29 more per month. They say your information is used to serve targeted advertising, and includes any links you follow and search terms you enter.

1 of 227 comments (clear)

  1. Re:Please note: by SethJohnson · · Score: 5, Interesting
    I'm thinking you set up your browser with the foxyproxy plugin so all normal http requests for HTML are forwarded via socks proxy. The heavy lifting stuff goes straight out and back in through the open fiber connection. I doubt AT&T wants to parse video files. They want to see and modify the clear-text HTTP stuff.

    The weird thing is this type of traditional snooping will be defeated as more content providers are switching over to HTTPS. AT&T aren't technical dummies, so they know that. I'm wondering if their scheme doesn't require a special browser plugin that automates an MITM attack on https....

    What's weirdest of all is that until now, federal law has protected the ISPs from liability over the content they transmit:

    Section 512(a) protects service providers who are passive conduits from liability for copyright infringement, even if infringing traffic passes through their networks. In other words, provided the infringing material is being transmitted at the request of a third party to a designated recipient, is handled by an automated process without human intervention, is not modified in any way, and is only temporarily stored on the system, the service provider is not liable for the transmission.

    The rationale behind that statue was that ISPs can't be held accountable for copyright-infringing material going over their wires because filtering it would be too onerous. If AT&T sets up such a monitoring system, it pretty well defeats the claim they don't know what their subscribers are transmitting / receiving.